Security

Security at DhiTantra.

Legal and compliance workflows demand absolute, non-negotiable confidentiality. DhiTantra is engineered from the ground up so that tier-1 law firms, corporate legal departments, and enterprise organizations can deploy legal automation workflows on high-stakes, sensitive matters with the ironclad controls expected from modern infrastructure.

Encryption & Infrastructure

  • Industry-Standard Encryption at Rest (e.g., AES-256) for all stored documents and user data.
  • Modern Encrypted Transport Protocols (e.g., TLS 1.2+ / 1.3) for all API and web traffic.
  • Infrastructure & Residency: Core databases and hosting infrastructure are localized in secure Indian cloud regions. Where specialized third-party AI models are utilized, data is transmitted via secure, encrypted enterprise APIs with strict zero-retention and zero-training policies.

Identity & Access Management

  • Role-Based Access Control (RBAC) ensuring strict data governance.
  • Workspace Logical Isolation: Complete separation of user and enterprise environments.
  • Secure Authentication protocols with session timeouts and monitoring.

Enterprise AI Security

  • Prompt Isolation: Query inputs are isolated and never bleed across sessions.
  • Zero Cross-Customer Data Leakage: Architecture designed for multi-tenant physical security and single-tenant logical isolation.
  • Training Wall: No customer data, uploaded files, or workspace content is ever used for shared model training without explicit contractual consent.

Operational Security & Compliance

  • Comprehensive Audit Logging for enterprise administrators.
  • Automated encrypted backups with strict retention policies.
  • Continuous threat monitoring and incident response readiness.
  • Compliance Roadmap: Our security program continues to evolve alongside rigorous customer, regulatory, and DPDP requirements.

Reporting a Vulnerability

Email security@dhitantra.ai with a clear description and reproduction steps. Please give us reasonable time to address issues before public disclosure.